Clinical Characteristics and Prognosis of Pediatric Idiopathic Multicentric Castleman Disease

Abstract

Idiopathic multicentric Castleman disease (iMCD) is a rare lymphoproliferative disorder characterized by marked heterogeneity among patients, with severity ranging from mild to life-threatening [1]. Although cases have been documented across all age groups, the diagnosis of iMCD typically occurs in the fifth decade of life and is relatively uncommon in children [1].

A previous analysis of the ACCELERATE dataset revealed that 95% of patients with iMCD diagnosed before the age of 30 (n = 35) had severe disease at onset [2]. A subsequent subgroup analysis focusing on the pediatric cohort (n = 19) highlighted a higher prevalence of the thrombocytopenia, anasarca, fever, renal dysfunction, and organomegaly (TAFRO) subtype [3]. These findings raise important questions about whether the severity and dominance of the TAFRO subtype in children suggest a poorer prognosis compared to that of adults. However, existing studies lack comprehensive data on pediatric iMCD. In this study, we aim to characterize the clinical features, treatment strategies, and outcomes of individuals diagnosed with iMCD at or under the age of 18, thereby contributing to a comprehensive understanding of the disease in this patient group.

This retrospective cohort study was conducted at two tertiary hospitals: Peking Union Medical College Hospital and Beijing Children's Hospital. We enrolled patients aged ≤ 18 years who were diagnosed with iMCD based on the Castleman Disease Collaborative Network (CDCN) diagnostic consensus [4] between January 2012 and January 2024. Patients were further classified into three clinical phenotypes: TAFRO, not otherwise specified (NOS), and idiopathic plasmacytic lymphadenopathy (IPL) [5]. Severe iMCD was evaluated according to the CDCN criteria at diagnosis [6].

The primary outcome was to describe the baseline characteristics. Other study outcomes included treatment regimens and responses, as well as prognosis. Treatments were further categorized into two main strategies: continual suppressive therapies, encompassing interleukin (IL)-6 targeting regimens, myeloma-like regimens, and mTOR-targeting regimens; and pulse therapy strategies, which include lymphoma-like regimens. The evaluation of treatment response adhered to the CDCN response criteria. Follow-up data, extending up to September 1st, 2024, were sourced from medical record systems and through telephone communication. Overall survival and time-to-next treatment (TTNT) were determined. Additional details on study methods are available in the Data S1 section.

A total of 23 patients (16 males and seven females [ratio, 2.3:1]), with a median age of 12 years at diagnosis (range: 2–18 years), were included in this analysis (Table S1). While there was no statistically significant difference in the age distribution between sexes, all children diagnosed at or before the age of 10 were male (Figure 1A). Seven (30.4%) met the TAFRO criteria, eight (34.8%) met the NOS criteria, and eight (34.8%) met the IPL criteria. The median age at diagnosis for TAFRO was significantly younger than those for NOS and IPL (7 years for TAFRO, 12 for NOS, and 15 for IPL; Figure 1B).

FIGURE 1

Open in figure viewerPowerPoint

Higher frequency of TAFRO detected in younger children with iMCD. (A) Age at diagnosis by sex. (B) Age of patients with different iMCD subtypes. (C) Age of patients by severity classification. (D) Relationship between platelet count and age in iMCD as determined by Spearman's correlation (R = 0.6847; p = 0.0003). Shaded regions represent the normal range. ****p < 0.0001; ***p < 0.001; **p < 0.01; ns, not significant.

Nine patients (39.1%) had severe iMCD at diagnosis, with significantly lower age compared to that of non-severe cases (Figure 1C). Regarding renal function, one patient with TAFRO and one with NOS exhibited renal insufficiency, with an estimated glomerular filtration rate (eGFR) below 60 mL/min*1.73 m². Two patients with TAFRO and one with NOS also presented with hemophagocytic lymphohistiocytosis (HLH) at disease onset. Laboratory tests revealed significant systemic inflammation in all groups, with no difference in C-reactive protein levels among them but significantly lower erythrocyte sedimentation rates in the TAFRO group than in the IPL group (Table S2). Hemoglobin, albumin, and creatinine levels showed no significant differences between groups. Seventeen (73.9%) patients demonstrated platelet abnormalities, comprising four cases of thrombocytopenia and 13 cases of thrombocytosis. All instances of thrombocytopenia occurred in younger children, and a significant correlation between platelet count and patient age was identified (Figure 1D).

Twenty-one patients received treatment, while two underwent spontaneous remission without specific iMCD therapy (Table S3). Continuous suppressive therapies accounted for 76.2% (n = 16) of first-line treatments, with IL-6-targeting agents being the most common (n = 11, 52.4%; Figure S1). Despite the lack of a statistically significant difference in remission rates across treatment regimens, the continual suppressive therapy exhibited higher TTNT than pulse therapy (p = 0.033; Figure S2). Regarding the first-line remission rate, the TAFRO subtype exhibited a relatively lower rate (28.6% for TAFRO; 71.4% for NOS; and 85.7% for IPL; p = 0.056). Compared to IL-6-targeting therapies, TAFRO also exhibited a more favorable response to myeloma-like regimens with a response rate of 100% versus 50%. Among the three patients with TAFRO who did not respond to first-line IL-6-targeting therapies (Table S3), two achieved remission after switching to myeloma-like regimens, while one achieved remission after switching from siltuximab to tocilizumab. Across all groups, no deaths were reported during a median follow-up period of 37 months (ranging from 5 to 114 months).

Among the three patients presenting with concurrent HLH at onset, two were initially managed for iMCD and one for HLH (Table S4). The two iMCD-targeted patients achieved continuous remission for both iMCD and HLH with tocilizumab- and bortezomib-based regimens, respectively. In contrast, the HLH-targeted patient achieved temporary remission with the HLH-2004 regimen but relapsed 1 year later. Subsequent treatment with allogeneic hematopoietic stem cell transplantation (allo-HSCT) resulted in complete remission of both iMCD and HLH. Moreover, 2 years later, iMCD relapsed, and tocilizumab was administered as the third-line regimen, which once again led to remission. As of the last follow-up (progression-free survival = 10 months), the patient had not experienced relapses of either HLH or iMCD.

Compared to adult iMCD [1], our pediatric iMCD cohort exhibited a male preponderance, with all patients diagnosed at ≤ 10 years being male, suggesting a potential early-onset phenotype in males. Clinical subtypes were evenly distributed, but pediatric iMCD showed a higher prevalence of TAFRO, especially in younger children, compared to adults. In contrast, IPL was more common in older children after pubertal onset. The younger age at diagnosis among TAFRO may explain the higher incidence of severe disease and thrombocytopenia observed in younger patients.

Regarding treatment, IL-6 targeting therapy is recommended as first-line therapy if available. However, due to its limited accessibility, siltuximab was not available on the Chinese market until 2022, and tocilizumab was both expensive and not approved for the treatment of iMCD in China. Additionally, both medications require continuous intravenous infusion. Therefore, myeloma-like therapies were also commonly used as first-line treatments for pediatric iMCD in China. Notably, one patient in our cohort successfully discontinued treatment after 30-month thalidomide-based regimens and has remained in lasting remission for 70 months. Whether these myeloma-like regimens could offer opportunities for cure or long-time treatment discontinuation remains an area of ongoing research.

Additionally, we identified three iMCD cases (13.0%) where HLH co-occurred at onset, and all three underwent genetic testing without revealing any causative genetic mutations. HLH can be familial or acquired etiologies such as autoinflammatory and autoimmune disorders, infections, and malignancies. Treatment of the underlying disorders may reverse the acquired HLH, as was true in our cases. In the case where allo-HSCT was administered, both iMCD and HLH were effectively managed. However, the TTNT was relatively short in terms of treatment intensity, suggesting that continual suppressive therapy strategies with moderate treatment intensity may still be suitable for patients with concurrent HLH.

In our cohort, two cases exhibited spontaneous remission, and although some patients received up to third-line treatment, no deaths occurred during the study period. Two patients initially underwent continuous renal replacement therapy due to an eGFR below 60; however, their renal function rapidly improved following treatment for iMCD. These results indicate that the overall prognosis for pediatric iMCD is favorable, as supported by previously published cases of pediatric MCD (total sample size = 16, seen in Table S5), which reported no deaths within median follow-up periods ranging from 1.0 to 3.8 years. We recognize the limitations of our study, primarily the small sample size. Nevertheless, our cohort of 23 pediatric patients with iMCD is among the first to provide insights into this subgroup. Our study identifies several unique characteristics of pediatric iMCD, including (1) a male preponderance, (2) a relatively high proportion of TAFRO subtype and severe disease in younger children, (3) a tendency to develop concurrent HLH, (4) effective response to continuous suppressive therapy, and (5) an overall favorable prognosis. Future research, through larger, multi-center studies, is essential to better validate these characteristics and long-term outcomes in pediatric iMCD.